Occupancy Detection – Office Room Environmental Sensors Dataset

Overview

The Occupancy Detection dataset contains experimental data from an office room instrumented with environmental and CO₂ sensors, created to evaluate statistical learning models for detecting whether the room is occupied.[web:199]

Data Collection

• Three files (training and two test sets) with a total of 20,560 time-stamped instances sampled every minute.[web:199]
• Ground-truth occupancy (0 or 1) was obtained from time-stamped pictures taken every minute and manually labeled as occupied or not occupied.[web:199]

Variables

• date and time fields plus an integer id for each observation.[web:199]
• Temperature (°C), Humidity (%), Light (Lux), CO₂ (ppm), and HumidityRatio (kg water-vapor/kg air).[web:199]
• Occupancy: binary target label, 0 for not occupied and 1 for occupied.[web:199]

Use Cases

• Developing occupancy detection models for smart HVAC and building automation.[web:199]
• Evaluating classification algorithms on realistic indoor environmental sensor data.[web:199]

License

The UCI page specifies that this dataset is licensed under Creative Commons Attribution 4.0 International (CC BY 4.0), which permits sharing and adaptation with appropriate credit.[web:199]